Market Basket Analysis¶

HTML Customization¶

In [2]:
from IPython.core.display import HTML
HTML("""
<style>
.output_png {
    display: table-cell;
    text-align: center;
    vertical-align: middle;
    horizontal-align: middle;
}
h1,h2 {
    text-align: center;
    background-color: black;
    padding: 20px;
    margin: 0;
    color: yellow;
    font-family: ariel;
    border-radius: 80px
}

h3 {
    text-align: center;
    border-style: solid;
    border-width: 3px;
    padding: 12px;
    margin: 0;
    color: black;
    font-family: ariel;
    border-radius: 80px;
    border-color: gold;
}

body, p {
    font-family: ariel;
    font-size: 15px;
    color: charcoal;
}
div {
    font-size: 14px;
    margin: 0;

}

h4 {
    padding: 0px;
    margin: 0;
    font-family: ariel;
    color: purple;
}
</style>
""")
Out[2]:

Workflow¶

  1. Import Packages
  2. Loading dataset
  3. Data Pre-Processing
  4. Performing Exploratory Data Analysis
  5. Apriori Implementation
  6. Result Customization

Import Packages¶

In [7]:
pip install apyori

Collecting apyori
  Downloading apyori-1.1.2.tar.gz (8.6 kB)
  Preparing metadata (setup.py): started
  Preparing metadata (setup.py): finished with status 'done'
Building wheels for collected packages: apyori
  Building wheel for apyori (setup.py): started
  Building wheel for apyori (setup.py): finished with status 'done'
  Created wheel for apyori: filename=apyori-1.1.2-py3-none-any.whl size=5975 sha256=746ef9e3b4999a70d99965e039c56e88b9f8c6cdd89dcdc15e5b4664df5edd16
  Stored in directory: c:\users\user\appdata\local\pip\cache\wheels\84\45\a4\8ade6576f75410d8162c6da1de0aa9df56c16c711acab5a813
Successfully built apyori
Installing collected packages: apyori
Successfully installed apyori-1.1.2
Note: you may need to restart the kernel to use updated packages.
In [8]:
import numpy as np # linear algebra
import pandas as pd # Data pre-processing
import seaborn as sns # Required for plotting
import matplotlib.pyplot as plt # Required for plotting

Loading Dataset¶

In [202]:
df = pd.read_csv("Online_Retail.csv")
df.head(10)
Out[202]:
InvoiceNo StockCode Description Quantity InvoiceDate UnitPrice CustomerID Country
0 536365 85123A WHITE HANGING HEART T-LIGHT HOLDER 6 2010-12-01 08:26:00 2.55 17850.0 United Kingdom
1 536365 71053 WHITE METAL LANTERN 6 2010-12-01 08:26:00 3.39 17850.0 United Kingdom
2 536365 84406B CREAM CUPID HEARTS COAT HANGER 8 2010-12-01 08:26:00 2.75 17850.0 United Kingdom
3 536365 84029G KNITTED UNION FLAG HOT WATER BOTTLE 6 2010-12-01 08:26:00 3.39 17850.0 United Kingdom
4 536365 84029E RED WOOLLY HOTTIE WHITE HEART. 6 2010-12-01 08:26:00 3.39 17850.0 United Kingdom
5 536365 22752 SET 7 BABUSHKA NESTING BOXES 2 2010-12-01 08:26:00 7.65 17850.0 United Kingdom
6 536365 21730 GLASS STAR FROSTED T-LIGHT HOLDER 6 2010-12-01 08:26:00 4.25 17850.0 United Kingdom
7 536366 22633 HAND WARMER UNION JACK 6 2010-12-01 08:28:00 1.85 17850.0 United Kingdom
8 536366 22632 HAND WARMER RED POLKA DOT 6 2010-12-01 08:28:00 1.85 17850.0 United Kingdom
9 536367 84879 ASSORTED COLOUR BIRD ORNAMENT 32 2010-12-01 08:34:00 1.69 13047.0 United Kingdom
In [11]:
df.isnull().sum().sort_values(ascending=False)
Out[11]:
CustomerID     133600
Description      1454
InvoiceNo           0
StockCode           0
Quantity            0
InvoiceDate         0
UnitPrice           0
Country             0
dtype: int64

Data Pre-Processing¶

In [18]:
df[['CustomerID']] = df[['CustomerID']].apply(pd.to_numeric)
In [19]:
df["InvoiceDate"] = pd.to_datetime(df["InvoiceDate"])
In [20]:
df.info() # Checking data type information for validation purposes

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 495478 entries, 0 to 495477
Data columns (total 8 columns):
 #   Column       Non-Null Count   Dtype         
---  ------       --------------   -----         
 0   InvoiceNo    495478 non-null  object        
 1   StockCode    495478 non-null  object        
 2   Description  494024 non-null  object        
 3   Quantity     495478 non-null  int64         
 4   InvoiceDate  495478 non-null  datetime64[ns]
 5   UnitPrice    495478 non-null  float64       
 6   CustomerID   361878 non-null  float64       
 7   Country      495478 non-null  object        
dtypes: datetime64[ns](1), float64(2), int64(1), object(4)
memory usage: 30.2+ MB

Performing Exploratory Data Analysis¶

In [23]:
df.head()
Out[23]:
InvoiceNo StockCode Description Quantity InvoiceDate UnitPrice CustomerID Country
0 536365 85123A WHITE HANGING HEART T-LIGHT HOLDER 6 2010-12-01 08:26:00 2.55 17850.0 United Kingdom
1 536365 71053 WHITE METAL LANTERN 6 2010-12-01 08:26:00 3.39 17850.0 United Kingdom
2 536365 84406B CREAM CUPID HEARTS COAT HANGER 8 2010-12-01 08:26:00 2.75 17850.0 United Kingdom
3 536365 84029G KNITTED UNION FLAG HOT WATER BOTTLE 6 2010-12-01 08:26:00 3.39 17850.0 United Kingdom
4 536365 84029E RED WOOLLY HOTTIE WHITE HEART. 6 2010-12-01 08:26:00 3.39 17850.0 United Kingdom
In [26]:
df.groupby(['InvoiceDate'])['Description'].agg(['count']).plot(figsize=(12,5), grid=True, title="Total Number of Items Sold by Date").set(xlabel="Date", ylabel="Total Number of Items Sold")
Out[26]:
[Text(0.5, 0, 'Date'), Text(0, 0.5, 'Total Number of Items Sold')]
In [41]:
## Creating distribution of Item Sold
Item_distr = df.groupby(by = "Description").size().reset_index(name='Frequency').sort_values(by = 'Frequency',ascending=False).head(10)

## Declaring variables

bars = Item_distr["Description"]
height = Item_distr["Frequency"]
x_pos = np.arange(len(bars))

## Defining Figure Size

plt.figure(figsize=(16,9))

# Create bars
ax= plt.subplot() # to use in overlapping function
plt.bar(x_pos, height, color=(0.3, 0.4, 0.6, 0.6))

# Add title and axis names
plt.title("Top 10 Sold Items")
plt.xlabel("Item Name")
plt.ylabel("Number of Quantity Sold")

# Create names on the x-axis
plt.xticks(x_pos, bars)

# Function to avoid overlapping
plt.setp(ax.get_xticklabels(), rotation=30, ha='right')

# Show graph
plt.show()
In [197]:
# Month-Year Sales
df_date=df.set_index(['InvoiceDate']) ## Setting date as index for plotting purpose
df_date.head()
Out[197]:
InvoiceNo StockCode Description Quantity UnitPrice CustomerID Country
InvoiceDate
2010-12-01 08:26:00 536365 85123A WHITE HANGING HEART T-LIGHT HOLDER 6 2.55 17850.0 United Kingdom
2010-12-01 08:26:00 536365 71053 WHITE METAL LANTERN 6 3.39 17850.0 United Kingdom
2010-12-01 08:26:00 536365 84406B CREAM CUPID HEARTS COAT HANGER 8 2.75 17850.0 United Kingdom
2010-12-01 08:26:00 536365 84029G KNITTED UNION FLAG HOT WATER BOTTLE 6 3.39 17850.0 United Kingdom
2010-12-01 08:26:00 536365 84029E RED WOOLLY HOTTIE WHITE HEART. 6 3.39 17850.0 United Kingdom
In [44]:
df_date.groupby(['InvoiceDate'])['Description'].agg(['count']).plot(figsize=(12,5), grid=True, 
            title="Total Number of Items Sold by Date").set(xlabel="Date", ylabel="Total Number of Items Sold")
Out[44]:
[Text(0.5, 0, 'Date'), Text(0, 0.5, 'Total Number of Items Sold')]
In [45]:
df_date.resample("M")['Description'].count().plot(figsize = (20,8), grid = True, 
                title = "Number by Items Sold by Month").set(xlabel = "Date", ylabel = "Number of Items Sold")
Out[45]:
[Text(0.5, 0, 'Date'), Text(0, 0.5, 'Number of Items Sold')]
In [58]:
df_date["Hour"] = df_date.index.hour
df_date["Weekday"] = df_date.index.weekday + 1

df_date.head(10)
Out[58]:
InvoiceNo StockCode Description Quantity UnitPrice CustomerID Country Hour Weekday
InvoiceDate
2010-12-01 08:26:00 536365 85123A WHITE HANGING HEART T-LIGHT HOLDER 6 2.55 17850.0 United Kingdom 8 3
2010-12-01 08:26:00 536365 71053 WHITE METAL LANTERN 6 3.39 17850.0 United Kingdom 8 3
2010-12-01 08:26:00 536365 84406B CREAM CUPID HEARTS COAT HANGER 8 2.75 17850.0 United Kingdom 8 3
2010-12-01 08:26:00 536365 84029G KNITTED UNION FLAG HOT WATER BOTTLE 6 3.39 17850.0 United Kingdom 8 3
2010-12-01 08:26:00 536365 84029E RED WOOLLY HOTTIE WHITE HEART. 6 3.39 17850.0 United Kingdom 8 3
2010-12-01 08:26:00 536365 22752 SET 7 BABUSHKA NESTING BOXES 2 7.65 17850.0 United Kingdom 8 3
2010-12-01 08:26:00 536365 21730 GLASS STAR FROSTED T-LIGHT HOLDER 6 4.25 17850.0 United Kingdom 8 3
2010-12-01 08:28:00 536366 22633 HAND WARMER UNION JACK 6 1.85 17850.0 United Kingdom 8 3
2010-12-01 08:28:00 536366 22632 HAND WARMER RED POLKA DOT 6 1.85 17850.0 United Kingdom 8 3
2010-12-01 08:34:00 536367 84879 ASSORTED COLOUR BIRD ORNAMENT 32 1.69 13047.0 United Kingdom 8 3
In [60]:
df['Description'].value_counts()
Out[60]:
WHITE HANGING HEART T-LIGHT HOLDER     2271
JUMBO BAG RED RETROSPOT                2001
REGENCY CAKESTAND 3 TIER               1859
PARTY BUNTING                          1634
LUNCH BAG RED RETROSPOT                1460
                                       ... 
MIA                                       1
PINK CRYSTAL GUITAR PHONE CHARM           1
showroom                                  1
TRAVEL CARD WALLET DOTCOMGIFTSHOP         1
ORANGE PENDANT TRIPLE SHELL NECKLAC       1
Name: Description, Length: 4202, dtype: int64
In [192]:
#df['Description'].nunique() # There are 3950 buyers represented by their ID
#df['Description'].value_counts().nlargest(20) # Showing the mode but in the list
#Sorting top 20 items
df_top20 = df.groupby(by = "Description").size().reset_index(name='Frequency').sort_values(by = 'Frequency',ascending=False).head(20)
df_top20.head()
Out[192]:
Description Frequency
3898 WHITE HANGING HEART T-LIGHT HOLDER 2271
1858 JUMBO BAG RED RETROSPOT 2001
2899 REGENCY CAKESTAND 3 TIER 1859
2460 PARTY BUNTING 1634
2045 LUNCH BAG RED RETROSPOT 1460
In [64]:
import plotly.express as px
In [93]:
fig = px.bar(df,
       x = df_top20['Description'], 
       y = df_top20['Frequency'],
       color= df_top20['Description'] )
    
fig.update_layout(
xaxis_title= 'Description',
yaxis_title="Count",
legend_title= 'Description',
font_family="Courier New",
font_color="blue",
title_font_family="Times New Roman",
title_font_color="red",
legend_title_font_color="green")

    
fig.show()
In [159]:
df_2 = df.groupby(['CustomerID','InvoiceDate'])['Description'].apply(sum)
In [164]:
transactions = [a[1]['Description'].tolist() for a in list(df.groupby(['CustomerID']))]

C:\Users\USER\AppData\Local\Temp\ipykernel_6536\773814388.py:1: FutureWarning:

In a future version of pandas, a length 1 tuple will be returned when iterating over a groupby with a grouper equal to a list of length 1. Don't supply a list with a single grouper to avoid this warning.
In [166]:
from mlxtend.preprocessing import TransactionEncoder
In [167]:
te = TransactionEncoder()
te_ary = te.fit(transactions).transform(transactions)
In [170]:
transactions = pd.DataFrame(te_ary, columns=te.columns_)
pf = transactions.describe()
pf
Out[170]:
4 PURPLE FLOCK DINNER CANDLES 50'S CHRISTMAS GIFT BAG LARGE DOLLY GIRL BEAKER I LOVE LONDON MINI BACKPACK NINE DRAWER OFFICE TIDY OVAL WALL MIRROR DIAMANTE RED SPOT GIFT BAG LARGE SET 2 TEA TOWELS I LOVE LONDON SPACEBOY BABY GIFT SET TOADSTOOL BEDSIDE LIGHT ... ZINC STAR T-LIGHT HOLDER ZINC SWEETHEART SOAP DISH ZINC SWEETHEART WIRE LETTER RACK ZINC T-LIGHT HOLDER STAR LARGE ZINC T-LIGHT HOLDER STARS LARGE ZINC T-LIGHT HOLDER STARS SMALL ZINC TOP 2 DOOR WOODEN SHELF ZINC WILLIE WINKIE CANDLE STICK ZINC WIRE KITCHEN ORGANISER ZINC WIRE SWEETHEART LETTER TRAY
count 3950 3950 3950 3950 3950 3950 3950 3950 3950 3950 ... 3950 3950 3950 3950 3950 3950 3950 3950 3950 3950
unique 2 2 2 2 2 2 2 2 2 2 ... 2 2 2 2 2 2 2 2 2 2
top False False False False False False False False False False ... False False False False False False False False False False
freq 3920 3854 3872 3904 3926 3911 3874 3834 3874 3949 ... 3949 3932 3885 3835 3948 3782 3940 3827 3939 3933

4 rows × 3860 columns

In [171]:
pf.iloc[0]-pf.iloc[3]
Out[171]:
 4 PURPLE FLOCK DINNER CANDLES      30 
 50'S CHRISTMAS GIFT BAG LARGE      96 
 DOLLY GIRL BEAKER                  78 
 I LOVE LONDON MINI BACKPACK        46 
 NINE DRAWER OFFICE TIDY            24 
                                    .. 
ZINC T-LIGHT HOLDER STARS SMALL     168
ZINC TOP  2 DOOR WOODEN SHELF       10 
ZINC WILLIE WINKIE  CANDLE STICK    123
ZINC WIRE KITCHEN ORGANISER         11 
ZINC WIRE SWEETHEART LETTER TRAY    17 
Length: 3860, dtype: object
In [194]:
f = pf.iloc[0]-pf.iloc[3]
a = f.tolist()
b = list(f.index)
item = pd.DataFrame([[a[r],b[r]]for r in range(len(a))], columns=['Count','Item'])
item = item.sort_values(['Count'], ascending=False).head(50)
#transactions
In [195]:
item.head()
Out[195]:
Count Item
3681 823 WHITE HANGING HEART T-LIGHT HOLDER
2751 773 REGENCY CAKESTAND 3 TIER
2334 659 PARTY BUNTING
214 643 ASSORTED COLOUR BIRD ORNAMENT
2302 596 PAPER CHAIN KIT 50'S CHRISTMAS
In [174]:
import matplotlib.pyplot as plt
import seaborn as sns

from wordcloud import WordCloud

plt.rcParams['figure.figsize'] = (15, 15)
wordcloud = WordCloud(background_color = 'white', width = 1200,  height = 1200, max_words = 121).generate(str(item['Item']))
plt.imshow(wordcloud)
plt.axis('off')
plt.title('Items',fontsize = 20)
plt.show()

C:\Users\USER\anaconda3\lib\site-packages\wordcloud\wordcloud.py:106: MatplotlibDeprecationWarning:

The get_cmap function was deprecated in Matplotlib 3.7 and will be removed two minor releases later. Use ``matplotlib.colormaps[name]`` or ``matplotlib.colormaps.get_cmap(obj)`` instead.
In [175]:
fig = px.treemap(item, path=['Item'], values='Count')
fig.show()

C:\Users\USER\anaconda3\lib\site-packages\plotly\io\_renderers.py:395: DeprecationWarning:

distutils Version classes are deprecated. Use packaging.version instead.

C:\Users\USER\anaconda3\lib\site-packages\plotly\io\_renderers.py:395: DeprecationWarning:

distutils Version classes are deprecated. Use packaging.version instead.

Apriori Implementation¶

In [154]:
from mlxtend.frequent_patterns import apriori, association_rules
In [196]:
cust_level = df[["CustomerID", "Description"]].sort_values(by = "CustomerID", ascending = False) ## Selecting only required variables for modelling
cust_level['Description'] = cust_level['Description'].str.strip() # Removing white spaces if any
cust_level.head()
Out[196]:
CustomerID Description
357882 18287.0 CHRISTMAS TREE HANGING SILVER
183242 18287.0 SMALL PURPLE BABUSHKA NOTEBOOK
183244 18287.0 SMALL YELLOW BABUSHKA NOTEBOOK
183245 18287.0 LARGE YELLOW BABUSHKA NOTEBOOK
183246 18287.0 LARGE PURPLE BABUSHKA NOTEBOOK
In [177]:
transactions = [a[1]['Description'].tolist() for a in list(cust_level.groupby(['CustomerID']))] ## Combing all the items in list format for each cutomer

C:\Users\USER\AppData\Local\Temp\ipykernel_6536\1472369653.py:1: FutureWarning:

In a future version of pandas, a length 1 tuple will be returned when iterating over a groupby with a grouper equal to a list of length 1. Don't supply a list with a single grouper to avoid this warning.
In [198]:
from apyori import apriori ## Importing apriori package
rules = apriori(transactions = transactions, min_support = 0.005, min_confidence = 0.08, min_lift = 3, min_length = 2, max_length = 2) ## Model Creation
results = list(rules) ## Storing results in list format for better visualisation

Results Customization¶

In [183]:
## Creating user-defined function for arranging the results obtained from model into readable format
def inspect(results):
    lhs         = [tuple(result[2][0][0])[0] for result in results]
    rhs         = [tuple(result[2][0][1])[0] for result in results]
    supports    = [result[1] for result in results]
    confidences = [result[2][0][2] for result in results]
    lifts       = [result[2][0][3] for result in results]
    return list(zip(lhs, rhs, supports, confidences, lifts))
resultsinDataFrame = pd.DataFrame(inspect(results), columns = ['Left Hand Side', 'Right Hand Side', 'Support', 'Confidence', 'Lift'])
In [184]:
resultsinDataFrame.nlargest(n=10, columns="Lift") ## Showing best possible scenarios
Out[184]:
Left Hand Side Right Hand Side Support Confidence Lift
47076 PANTRY HOOK BALLOON WHISK PANTRY HOOK TEA STRAINER 0.005316 0.840000 138.250000
48084 PARTY PIZZA DISH BLUE POLKADOT PARTY PIZZA DISH PINK POLKADOT 0.005570 0.880000 128.740741
44148 MIRRORED WALL ART GENTS MIRRORED WALL ART LADIES 0.005823 0.793103 125.310345
23630 ENAMEL PINK COFFEE CONTAINER ENAMEL PINK TEA CONTAINER 0.005823 0.766667 108.154762
8148 BICYCLE SAFTEY WALL ART DOG LICENCE WALL ART 0.005570 0.611111 92.841880
10060 BLUE VICTORIAN FABRIC OVAL BOX RED VICTORIAN FABRIC OVAL BOX 0.005823 0.821429 90.128968
44811 ORIGAMI ROSE INCENSE/CANDLE SET ORIGAMI SANDLEWOOD INCENSE/CAND SET 0.005823 0.718750 88.720703
57491 SILK PURSE BABUSHKA BLUE SILK PURSE BABUSHKA RED 0.005316 0.636364 83.787879
18950 COFFEE MUG BLUE PAISLEY DESIGN COFFEE MUG PINK PAISLEY DESIGN 0.005823 0.638889 81.406810
48085 PARTY PIZZA DISH PINK POLKADOT PARTY PIZZA DISH RED RETROSPOT 0.005063 0.740741 79.079079
In [199]:
fig=px.scatter(resultsinDataFrame['Support'], resultsinDataFrame['Confidence'])
fig.update_layout(
    xaxis_title="support",
    yaxis_title="confidence",
   
    font_family="Courier New",
    font_color="blue",
    title_font_family="Times New Roman",
    title_font_color="red",
    title=('Support vs Confidence')
    
)

fig.show()

C:\Users\USER\anaconda3\lib\site-packages\plotly\io\_renderers.py:395: DeprecationWarning:

distutils Version classes are deprecated. Use packaging.version instead.

C:\Users\USER\anaconda3\lib\site-packages\plotly\io\_renderers.py:395: DeprecationWarning:

distutils Version classes are deprecated. Use packaging.version instead.
In [200]:
fig=px.scatter(resultsinDataFrame['Support'], resultsinDataFrame['Lift'])
fig.update_layout(
    xaxis_title="support",
    yaxis_title="lift",
   
    font_family="Courier New",
    font_color="blue",
    title_font_family="Times New Roman",
    title_font_color="red",
    title=('Support vs Confidence')
    
)

fig.show()

C:\Users\USER\anaconda3\lib\site-packages\plotly\io\_renderers.py:395: DeprecationWarning:

distutils Version classes are deprecated. Use packaging.version instead.

C:\Users\USER\anaconda3\lib\site-packages\plotly\io\_renderers.py:395: DeprecationWarning:

distutils Version classes are deprecated. Use packaging.version instead.
In [201]:
fit = np.polyfit(resultsinDataFrame['Lift'], resultsinDataFrame['Confidence'], 1)
fit_fn = np.poly1d(fit)
plt.plot(resultsinDataFrame['Lift'], resultsinDataFrame['Confidence'], 'yo', resultsinDataFrame['Lift'], 
fit_fn(resultsinDataFrame['Lift']))
plt.xlabel('Lift')
plt.ylabel('Confidence')
plt.title('lift vs Confidence')
Out[201]:
Text(0.5, 1.0, 'lift vs Confidence')

THANK YOU¶